runtime-fallback-graph fails on stage 1 .NET 8 builds

[tmds]

This test fails on a .NET 8 unknown distro build.

This is not unexpected since the runtime logic that updates the runtime graph during the build was removed.

Probably a stage 2 build will pass the test, because the rid is known by the final output of the build.
This should be verified.

Because the logic around rids has seen a major overhaul with .NET 8, we should verify how the issue that this test is guarding against is behaving with the stage 1 and stage 2 outputs.

With some recent changes (SDK switching from legacy graph to a trimmed graph), it should become so that all stage 1 builds from a Microsoft SDK are in the "previously unknown distro" case.

cc @omajid

[tmds]

Because the logic around rids has seen a major overhaul with .NET 8, we should verify how the issue that this test is guarding against is behaving with the stage 1 and stage 2 outputs.

The information added to the runtimes section should only be used when an application explicitly opts into it.
By default, a hard coded list from the host is used instead.

[omajid]

Any reason this wouldn't also apply to .NET 9?

[tmds]

I don't think anyone cares about fixing this, so I expect it to continue behaving like this.